Sender Rotation Plan (MCP)
Build a data-driven sender rotation plan for any workspace in seconds—using the Outreach Magic MCP inside Claude or Cursor.
Instead of spending an hour in a spreadsheet calculating scores, domain caps, and warmup health, you ask the MCP one question and get a tiered pick list: who's in, who's backup, who's out—and why.
What you get
- • Tier A — top performers by reply rate, bounce rate, warmup health, and domain capacity
- • Tier B — solid senders with smaller sample sizes or slightly lower scores
- • Backup — available but deprioritized
- • Excluded — with exact reasons (low warmup, bounce spike, stale sync, domain cap hit, etc.)
The plan factors in:
- • Sends, replies, and bounces over your lookback window
- • Warmup score and daily send limits
- • Domain capacity buckets (how many senders per domain can stay in rotation)
- • Peer comparison across domains at similar send volume
- • Workspace-level baselines so senders are compared against relevant peers
Prerequisites
- • A Pro workspace (sender health sync is Pro-only)
- • MCP enabled for your workspace
- • Claude Desktop, Cursor, or Windsurf connected to your Outreach Magic MCP
Step 1: Enable MCP for your workspace
- 1Go to Settings in your Outreach Magic workspace.
- 2Open the MCP tab.
- 3Copy your MCP token—you'll use it in Step 2.
Step 2: Add the MCP to Claude or Cursor
Claude (web or desktop)
Go to Settings → Connectors → Add custom connector and paste your MCP URL:
https://mcp.outreachmagic.io?token=YOUR_MCP_TOKENClick Add. No config file editing required.
Cursor
Open your Cursor MCP config and add the following under mcpServers (merge with any servers you already have). Replace om-your-workspace with a short label for this workspace (e.g. om-clienta) and substitute your real token for YOUR_MCP_TOKEN.
{
"mcpServers": {
"om-your-workspace": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.outreachmagic.io?token=YOUR_MCP_TOKEN"
]
}
}
}Save and reload Cursor.
Step 3: Run the rotation plan query
Once the MCP is connected, open a new conversation in Claude or Cursor and ask:
Can you help me build a sender rotation plan for [workspace-name]?
Provide full details and exactly which sender accounts we should use.
Export it as a markdown file.Example
Can you help me build a sender rotation plan for om-clienta,
provide me with the full details and exactly which sender accounts
we should use in a md fileThe MCP will pull current sender health data, score each account, and return a tiered plan with reasoning.
What the output looks like
The rotation plan is sorted by tier (A → B → backup → excluded), then by score descending. Reply and bounce rates reflect the lookback window.
| Tier | Sender | Score | Reply rate | Bounce rate | Warmup | Daily limit |
|---|---|---|---|---|---|---|
| A | lila@… | 57.96 | 62.50% | 0.00% | 100.0 | 2 |
| A | rose@… | 49.22 | 50.00% | 0.00% | 100.0 | 2 |
| A | pax@… | 40.49 | 37.50% | 0.00% | 100.0 | 2 |
| B | clar@… | 23.02 | 12.50% | 0.00% | 100.0 | 2 |
Excluded senders include the reason: low warmup, bounce spike, domain cap, stale sync, or provider trend.
How scoring works
Each sender gets a composite score based on:
- • Reply weight — heavier than other signals; configurable
- • Bounce penalty — inverse score; spike = immediate flag
- • Warmup health — score below threshold gets deprioritized
- • Sample size adjustment — small-volume senders get a confidence discount
- • Domain capacity — caps on how many accounts per domain can be Tier A
- • Peer comparison — optional penalty when a domain has volume but zero replies while similar domains reply actively
You can ask the MCP to adjust weights or lookback window in your query.
Audit mode
For a full breakdown with reasoning per sender, add to your query:
Include audit mode with full scoring detail per sender.This surfaces the intermediate score components—useful for QA or explaining rotation decisions to a client.
Tips
Run weekly. Sender health changes; last week's Tier A may have warmed down.
Domain summaries. Ask for per-domain rollups if you want aggregate performance before picking accounts.
New workspaces. If sample sizes are small, ask for a conservative confidence threshold so low-data senders aren't over-promoted.
Fresh sync. The plan works best when sender health sync is current. If data looks stale, check the sync timestamp in workspace settings.
Related
- Configure Senders — UI guide for sender accounts and workspaces
- Features — Sender & domain health overview (Pro)
- API Reference — REST API hub and automation endpoints
Ready to connect MCP?
Open your workspace, copy your MCP token from Settings, and run your first rotation plan in Claude or Cursor.